This site provides a list of sensors of various types that generate streams of data. In some cases, for example with accelerometers, the data stream will be fast, large and complex. Even with simple data such as temperature readings, making business decisions based on the data can be complex. In all cases, there’s a stream of data that requires Time Series Data (TSD) analysis.
There’s a new article at IoT World Today explaining how TSD analysis is a new frontier. Developers need to learn data science and in some cases AI deep learning to make sense of the data.
The cost of adding sensors is low, and the required infrastructure is now ubiquitous. The major challenge is the availability of expertise to realize value from the data streams
Care needs to be taken to collect data at the right rate, fast enough to capture the required features in the data but not too fast so as to overwhelm storage and processing. When there’s too much data, edge processing needs to be considered which also allows for faster alerts and notifications than if the data is processed at a server.
There’s also sometimes need to mix and match data from different sensors for analysis and display. There often needs to be a pragmatic balance between accuracy and speed of analysis.
With so much more data to digest, the bottleneck in the process really becomes the engineers and computer and data scientists that used to devise, test and implement algorithms largely by hand-using high-level languages
IoTBusinessNews has an article, based on Berg Insight research, saying 3.4 million cellular IoT gateways were shipped globally during 2018 that’s growing at a compound annual growth rate (CAGR) of 18.5 percent to reach annual reach 8.0 million in 2023.
Several players continue to have inorganic growth as a strategy and further consolidation can be expected among IoT gateway vendors in 2019–2020
Utilities and those in Primary Processing as well as Industrial/Manufacturing and Transportation and logistics for track and trace/asset management functions, will continue to provide the largest markets for unlicensed LPWA in the future.
Advanced Data Path (ADP) technology allows noise filtering and sensor signal processing normally carried out by the MCU to be performed by the accelerometer.
This means the board containing the Accelerometer does less work and uses less power. It also enables the sensing frequency and acceleration detection to have wide ranges of 8,500Hz and ±64g respectively making this device suitable for industrial applications.
Fierce Electronics has an article mentioning LoRa sensors and gateways will be used aboard ocean freighters. Wilhelmsen will implement a large IoT platform aboard ocean-going freighters worldwide using LoRa.
“LoRa was the right choice of wireless technology, since its range is long enough to service an entire ship but not so long-range that it uses too much power that can quickly exhaust batteries in sensors”
A typical freighter will have 2,000 or more sensors sending data to three gateways.
The premise is to use consumer smartphones as the medium for connectivity. The Nodle Cash app gives users cash in return for IoT connectivity. The app uses Bluetooth to connect to manufacturers’ IoT sensors.
This solution is not for real-time connectivity but instead for scenarios where sensor update periods aren’t critical. The solution relies on there being enough consumer smartphone installs across the required geographic coverage.
Nodle has already raised $3.5 million in seed funding.